Eric Harman: Old project, New Post: Light Tracker

AKA Evil Eye

Project #2

Forward: Heliotropism in natural systems has evolved over millennia as a means of regulating energy capture and utilization. The mechanisms by which natural systems (e.g. sunflowers) implement solar tracking are complex, but robust (perhaps due to the built-in redundancy). In this exercise you will implement artificial systems that can track a moving light source, using your Arduino kits and some additional hardware. This exercise will serve as a foundation for more complex mechanisms you will develop under the theme of heliotropic smartsurfaces.

Your team's task is to design, build, program and test a device that tracks a light source (a handheld flashlight). This system should operate on dual axes and be active. Ideally it should be capable of:

Tracking the light horizontally
Tracking the light vertically
Indicating when it is in alignment with the light.

Initially, we were asked to accomplish this task using only cardboard and any other materials we could find in DL1 (DesignLab 1) and complete the task within a matter of hours. It was daunting to say the very least. Attempting to create motion, and keep it accurate and repeatable, was was completely at odds with the tools(utility blade and ruler) we had at our disposal within our limited time frame. After an initial Brainstorm session, the group decided we could be most effective if we split into subgroups; programming and construction.

We would then reconvene to build the circuit based

on physical constraints of the model as well as the necessity of our program.

Before breaking into sub-groups we did a little sketching and brainstorming on "how" we thought our tracker should behave in terms of navigating space. The image below shows this process.

The rough sketches depict the following:

Our Perception of how an "eye" moves to find a point in space
An observatory as a known system of application
The employment of some sort of barrier or wall to cast shadows across 4 LDR's. We thought this might be a good mechanical approach to determining if the tracker is normal to the light source, in tandem to our programming approach.

By the end of class, our group was unable to produce a functioning model. To our pleasure, the class was notified that we were to continue working and tweaking our Trackers for the next the class. Hooray!

For the next week, Mat and myself worked to construct a laser-cut model that could have all the electronics attached by the group at our next meeting. We chose the acrylic for its strength as well as ease of fabricating and availability. We used Black FoamCore in the areas that needed to be curved or in the case of the LDR divider, we needed a non-reflective surface.

This design functioned and fulfilled the requirements adequately. That being said, we realized the following issues:

The Acrylic needed a more rigid mechanical attachment to the servo. We utilized a Methalene-Chloride Based solvent to glue a servo connector to the axis. During Critique, this joint failed on the top axis.
Accounting for servo limitations. Travel was limited to 180 degrees of rotation (Our group mistakenly thought it was 360). Thus our model was not truly able to navigate a full hemisphere of travel.
Red was not a flattering color for our model. The Tracking indicator LED glowed an ominous red color and our group thought that it was a little creepy. Personally, I felt it looked like something out of a "Terminator" movie...